基于马尔可夫链的城市景观结构演化模拟及预测

Simulation and Prediction of Urban Landscape Structure Evolution Based on Markov Model:A Case Study of Yiwu City

以1996年、1999年、2002年和2005年遥感影像为数据源,对义乌城市景观演化进行研究。结果表明,农业景观是研究区景观演化最重要的"源",其次是水体景观和林地景观;其他景观都是城市景观演化的"汇"。利用马尔科夫模型,对义乌城市景观演化进行预测,结果显示,至2020年研究区农业景观仅剩641.09hm2,是1996年的10.04%,研究区几乎被人工景观所覆盖。

The process of urbanization is changing the structure and function of urban land- scape, and the purpose of research on dynamic variation and prediction of urban landscape is offering decision for urban development process design and landscape overall planning. Taking the remote sensing images of the year of 1996, 1999, 2002 and 2005 as data sources, this paper conducts some research on urban landscape structure evolution of Yiwu. The results show that agricultural landscape is the most important ＂source＂, followed by water land- scape and woodland landscape, while other landscapes are the ＂sinks＂ in this urban landscape evolution. Using of the Markov model, the result of prediction shows that the area of agri- cultural landscape will be only 641.09 hm2 in 2020, which is the 10.04% of the year of 1996. The study area will be almost covered by the artificial landscapes.